Neuroblastoma is the most common extracranial solid tumor in children. Disease remission can frequently be achieved in patients with high-risk neuroblastoma using surgery, radiation, and chemotherapy. Unfortunately, relapse is common, even many years later, and the majority of patients ultimately die of disease. The emerging stem cell theory of cancer suggests that cancer relapse may be caused by a rare tumorigenic progenitor cell that is resistant to conventional therapy. If true for neuroblastoma, improved cure rates will likely only be achieved by identification and therapeutic targeting of the neuroblastoma stem cell. Our overarching hypotheses are that (1) neuroblastomas arise from a cancer stem cell that can be isolated, purified, and characterized and (2) these cells can be specifically included along with bulk cancer cells for destruction using transcriptionally regulated oncolytic herpes viruses. Using previously described properties of neural stem cells, we identified and isolated cell subpopulations in human neuroblastoma cell cultures suggestive of a neuroblastoma stem cell. Cells express the neural stem cell marker CD133, exhibit a calcium channel-dependent side population, form spheres in serum-free media (that exhibit properties shared with bona fide neural stem cell-derived neurospheres), show evidence of asymmetric cell division, and are relatively resistant to the cytotoxic chemotherapy drug doxorubicin. Some of these cell populations are enriched for the neuroblastoma stem cell, demonstrated by their increased efficiency in forming tumors in mice at low cell inoculums. We propose to further refine the identification and isolation of the neuroblastoma stem cell from both cell lines and primary tumor samples using a combination of these characteristics (aim 1). Rather than rely exclusively on marker expression, this function-based, multi-parameter selection is unique, and should allow us to more precisely define the neuroblastoma cancer stem cell. We have also shown that sphere-forming cells express nestin, similar to bulk cells, and these cells are susceptible to virus-induced cytolysis by an oncolytic virus containing a critical virus gene under control of the nestin promoter. We propose to determine whether this virus exhibits a more pronounced antitumor effect in neuroblastoma xenografts, and if such an effect is due to improved targeting of the neuroblastoma stem cells (aim 2). The first aim should advance the field of cancer stem cell biology by pioneering a multi-parameter selection schema, and should advance the field of neuroblastoma biology by more precisely defining the stem cell phenotype and genotype. The second aim should provide justification for further development of the nestin-targeted virus as a therapeutic for neuroblastoma. It will also help open a new avenue for potential therapeutic strategies that are cytotoxic for the bulk of tumor cells but also include destruction of neuroblastoma stem cells. PUBLIC HEALTH RELEVANCE: Aside from brain tumors, neuroblastoma is the most common solid tumor in children. We have partly isolated human neuroblastoma tumor initiating cells (cancer stem cells) that may be responsible for relapsed disease. Here we will further purify and characterize these cells from cell lines and primary human neuroblastomas. We will also determine if a new type of therapy in development, oncolytic herpes viruses, can be directed to kill neuroblastoma stem cells and give better treatment results in mouse tumor models. The results should not only deepen our understanding of the pathogenesis of neuroblastoma but will also be a cornerstone for the development of herpes viruses and other new experimental therapeutics that target neuroblastoma stem cells.